62
Circular Dichroism in Protein Analysis
characteristic and the easiest to iden-
tify with VCD. Its frequency is relatively
free of overlapping transitions from other
parts of the molecule. An
α
-helix has a
major positive peak at
1640 cm
1
.An
antiparallel
β
-sheet has a split amide I
0
band, with a major, sharp positive peak at
1615 to 1620 cm
1
and a weaker pos-
itive one at
1690 cm
1
.R
an
d
omc
o
i
l
conformations give a broad positive peak
at
1650 cm
1
and a positive trough at
1620 cm
1
, which is quite universal in
band shape and intensity in most oligo-
and polypeptides.
The amide II band is a mixture of
N
H deformation and C
Ns
tre
tch
.The
differences among
α
-helix,
β
-sheet, and
random coil are less clear than those
with
the
amide I
band.
The
distinct
feature of the amide II VCD for the
α
-
helix is a dominant negative band; for
β
-sheet one observes a negative couplet
and for random coil there is a weaker
intensity positive band. The amide III
band of an
α
-helix is mostly positive
in the region of
1350 to 1250 cm
1
,
and for
β
-sheet it is mostly negative in
thesamereg
ion
.Theam
ideI
I
Ibandi
s
more sequence dependent, which affects
its ultimate utility for peptide backbone
conformational analyses.
VCD provides complementary data to
electronic CD. The spectra are richer
in
information,
but
the
instrumenta-
tion is not as developed or as widely
used as that for electronic CD. The ro-
tational strengths of vibrational transi-
tions are much less intense than those
of electronic transitions, which makes
it a challenge to achieve the necessary
signal/noise
ratio.
The
technique
has
clear promise, but for the remainder of
our discussion, we shall focus on elec-
tronic CD.
4
Estimation of Secondary Structure Content
The three-dimensional conformation of a
protein can be speciFed by the dihedral an-
gles of the backbone. Hydrogen-bonding
patterns and particular values of dihedral
angles lead to regular, local elements of
structure, known as
secondary structure
.
Secondary structures can be grouped into
α
-helices,
β
-sheet,
β
-turn, and random
co
i
l
.A
smo
s
tp
ep
t
id
e
sandp
ro
t
e
in
sa
r
e
mixtures of these four basic secondary
structures, the overall mean residue el-
lipticity at a speciFc wavelength, [
θ(λ)
], is
the sum of contributions from each sec-
ondary structure. It can be calculated by
the equation:
[
θ(λ)
]
=
f
α
[
θ(λ)
]
α
+
f
β
[
θ(λ)
]
β
+
f
γ
[
θ(λ)
]
γ
+
f
R
[
θ(λ)
]
R
+
noise
(
6
)
which is based on the fraction of each
secondary structure type:
α
-helix (
f
α
),
β
-
sheet (
f
β
),
β
-turn (
f
γ
), and random coil (
f
R
).
As a Frst approximation, the CD spectrum
of a protein is the sum of the appropriate
percentages of each component spectrum,
but the relative arrangements of structural
units and motifs may also contribute to the
observed CD spectrum.
There are many computer programs
that can analyze CD spectra to estimate
protein secondary structure content. The
major methods are either algorithm-based
or rely on databases. The former includes
constrained least squares, normalized least
squares,
singular
value
deconvolution,
parameterized Ft and neural networks.
±requently used software packages are
SELCON3, CONTIN, and Varslc. Database
approaches are available in SELCON3,
CONTIN, and SELCON. Most of these
facilities can be found on the web site:
previous page 1382 Encyclopedia of Molecular Cell Biology and Molecular Medicine read online next page 1384 Encyclopedia of Molecular Cell Biology and Molecular Medicine read online Home Toggle text on/off